| A"""
+ soup = self.soup(markup)
+ assert (
+ "A"
+ == soup.body.decode()
+ )
+
+ def test_extraction(self):
+ """
+ Test that extraction does not destroy the tree.
+
+ https://bugs.launchpad.net/beautifulsoup/+bug/1782928
+ """
+
+ markup = """
+
+ hello
+"""
+ soup = self.soup(markup)
+ [s.extract() for s in soup("script")]
+ [s.extract() for s in soup("style")]
+
+ assert len(soup.find_all("p")) == 1
+
+ def test_empty_comment(self):
+ """
+ Test that empty comment does not break structure.
+
+ https://bugs.launchpad.net/beautifulsoup/+bug/1806598
+ """
+
+ markup = """
+
+
+
+
+
+"""
+ soup = self.soup(markup)
+ inputs = []
+ for form in soup.find_all("form"):
+ inputs.extend(form.find_all("input"))
+ assert len(inputs) == 1
+
+ def test_tracking_line_numbers(self):
+ # The html.parser TreeBuilder keeps track of line number and
+ # position of each element.
+ markup = "\n \n\n\ntext "
+ soup = self.soup(markup)
+ assert 2 == soup.p.sourceline
+ assert 5 == soup.p.sourcepos
+ assert "sourceline" == soup.p.find("sourceline").name
+
+ # You can deactivate this behavior.
+ soup = self.soup(markup, store_line_numbers=False)
+ assert None is soup.p.sourceline
+ assert None is soup.p.sourcepos
+
+ def test_special_string_containers(self):
+ # The html5lib tree builder doesn't support this standard feature,
+ # because there's no way of knowing, when a string is created,
+ # where in the tree it will eventually end up.
+ pass
+
+ def test_html5_attributes(self):
+ # The html5lib TreeBuilder can convert any entity named in
+ # the HTML5 spec to a sequence of Unicode characters, and
+ # convert those Unicode characters to a (potentially
+ # different) named entity on the way out.
+ #
+ # This is a copy of the same test from
+ # HTMLParserTreeBuilderSmokeTest. It's not in the superclass
+ # because the lxml HTML TreeBuilder _doesn't_ work this way.
+ for input_element, output_unicode, output_element in (
+ ("⇄", "\u21c4", b"⇄"),
+ ("⊧", "\u22a7", b"⊧"),
+ ("𝔑", "\U0001d511", b"𝔑"),
+ ("≧̸", "\u2267\u0338", b"≧̸"),
+ ("¬", "\xac", b"¬"),
+ ("⫬", "\u2aec", b"⫬"),
+ (""", '"', b'"'),
+ ("∴", "\u2234", b"∴"),
+ ("∴", "\u2234", b"∴"),
+ ("∴", "\u2234", b"∴"),
+ ("fj", "fj", b"fj"),
+ ("⊔", "\u2294", b"⊔"),
+ ("⊔︀", "\u2294\ufe00", b"⊔︀"),
+ ("'", "'", b"'"),
+ ("|", "|", b"|"),
+ ):
+ markup = "%s " % input_element
+ div = self.soup(markup).div
+ without_element = div.encode()
+ expect = b"%s " % output_unicode.encode("utf8")
+ assert without_element == expect
+
+ with_element = div.encode(formatter="html")
+ expect = b"%s " % output_element
+ assert with_element == expect
+
+ @pytest.mark.parametrize(
+ "name,value",
+ [("document_declared_encoding", "utf8"), ("exclude_encodings", ["utf8"])],
+ )
+ def test_prepare_markup_warnings(self, name, value):
+ # html5lib doesn't support a couple of the common arguments to
+ # prepare_markup.
+ builder = self.default_builder()
+ kwargs = {name: value}
+ with warnings.catch_warnings(record=True) as w:
+ list(builder.prepare_markup("a", **kwargs))
+ [warning] = w
+ msg = str(warning.message)
+ assert (
+ msg
+ == f"You provided a value for {name}, but the html5lib tree builder doesn't support {name}."
+ )
+
+ def test_doctype_filtered(self):
+ # Since the html5lib parser doesn't support parse_only, this standard
+ # smoke-test test can't be run.
+ pass
--
cgit v1.2.3
|